Context & Data
Mexico’s Procurement System
Public contracts in Mexico are allocated through three mechanisms, governed by federal procurement law:
The law allows exceptions that permit invitation and direct allocation for contracts below certain value thresholds or during emergencies. In practice, approximately 70% of contract value was already being allocated through discretionary mechanisms before the pandemic, a baseline that suggests these exceptions are widely used.
The DCVB Measure
The paper’s main outcome is the Discrete-Contracts-Value-to-Budget (DCVB) ratio: the share of total contract value assigned through invitation auctions or direct allocation in a given institution-month.
DCVB is a risk measure, not a corruption measure. A high DCVB does not prove fraud or bribery. It indicates that a large share of public money bypassed competitive processes, which is consistent with legitimate emergency procurement but also with favoritism. The paper is explicit that it measures risk, and the limitations section discusses this distinction in detail.
Data
The data come from CompraNet, Mexico’s public procurement information system, systematized by the Instituto Mexicano para la Competitividad (IMCO). The dataset includes all types of public contracts: acquisitions, leases, public works, and services.
Sample construction: To obtain a balanced panel, only institutions that allocated at least one contract every month from January 2018 through December 2020 are retained. This yields:
- 64 federal institutions, covering approximately 75% of all federal public acquisition allocations
- 378,000 public acquisitions
- 2,304 observations at the institution-month level (64 institutions × 12 months × 3 years)
Descriptive Statistics
The table below (Table 1 in the paper) shows summary statistics at the institution-month level.
| Variable | Overall | Health Institutions | Non-Health Institutions |
|---|---|---|---|
| Average contract value (millions MXN) | 5.06 | 2.01 | 5.84 |
| Number of contracts per month | 164.46 | 402.48 | 103.79 |
| DCVB ratio | 0.700 | 0.665 | 0.709 |
| Observations | 2,304 | 468 | 1,836 |
Source: IMCO. Data are aggregated at the institution-month level.
Key baseline fact: Even before the pandemic, 70% of contract value was assigned through discretionary mechanisms. Healthcare institutions were slightly less discretionary than non-healthcare ones (66.5% vs 70.9%), which already hints at the pattern the paper will find post-pandemic.
The Healthcare vs. Non-Healthcare Split
The paper distinguishes two groups of institutions:
Health institutions — federal hospitals, the social security institute (IMSS), and health secretariats. These institutions were under intense public scrutiny during the pandemic.
Non-health institutions — institutions spanning defense, education, infrastructure, energy, and other sectors. These received comparatively less media attention during the pandemic.
This distinction is central to the paper’s main finding: while the public debate focused on healthcare procurement, the data tells a different story.
The COVID-19 Timeline in Mexico
The paper covers the period March to December 2020 as the treatment window. Key dates:
| Date | Event |
|---|---|
| March 2020 | COVID-19 declared a national health emergency. Lockdown begins. |
| April 2020 | Economic activity falls sharply. Stay-at-home orders in effect. |
| May 2020 | Stay-at-home order lifted. Economic activity partially resumes. |
The stay-at-home order ending in May 2020 corresponds to the moment when the event study shows the sharpest increase in the DCVB ratio.